Financial SponsorsPlatinum Sponsor
Many problems of interest in wireless communication, for example, sub-carrier and power allocation, OFDMA capacity, mobile-basestation association, antenna selection, etc. are combinatorial problems that are NP-hard to solve. In this tutorial, we will discuss techniques to get approximate solutions for these 'hard' combinatorial problems with bounded distance from the optimal solution. The main theme will be to exploit the sub-modularity of the corresponding objective functions, for which greedy algorithms can be shown to be close to the optimal solution.
Prof. Rahul Vaze obtained his Ph.D. from The University of Texas at Austin in 2009. Since Oct. 2009 he is a Reader at the School of Technology and Computer Science, Tata Institute of Fundamental Research, Mumbai, India. His research interest are in multiple antenna communication, ad hoc networks, combinatorial resource allocation. He is a co-recipient of the EURASIP best paper award for year 2010 for the Journal of Wireless Communication and Networking, and recipient of Indian National Science Academy's young scientist award for the year 2013, Indian National Academy of Engineering's young engineer award for the year 2013, Ramanath Cowsik medal from TIFR for the year 2014.
Motivated by the smart-dust paradigm, consider a physical field sampling setup where many precision-limited sensors have to acquire a spatial field in a possibly noisy environment. This can be also termed as a distributed field acquisition problem. In a centralized setup, tradition dictates that a smooth signal should be acquired using Nyquist style sampling. To avoid aliasing or to bandlimit a signal, an anti-aliasing prefilter can be used. To combat noise, the signal can be filtered to its (essential) bandwidth and then sampled. However, in a distributed field sampling setup with fixed sensors, lowpass anti-aliasing prefilter cannot be used. This is a fundamental limitation in spatial field sampling, and it unfurls a wide range of problems involving interplay of noise, oversampling, quantization, and aliasing. In this tutorial, the acquisition of bandlimited fields or smooth fields will be examined in the noiseless as well as noisy setting. In these paradigms, oversampling is expected to overcome the lack of ADC precision, lack of knowledge about sensor-location, as well as the effect of noise. For various classes of spatial fields, the tradeoffs between oversampling, ADC precision, distortion with respect to some chosen metric, and impact of noise will be examined. Special attention will be given to single-bit quantization as it captures the coarsest precision available with any (low-precision) sensor. The tutorial will assume minimal background on single-bit quantization and spatial acquisition.
Prof. Animesh Kumar received his BTech degree in 2001 in Electrical Engineering from Indian Institute of Technology Kanpur (India), and MS and PhD degrees in Electrical Engineering and Computer Science from University of California, Berkeley, CA in 2003 and 2008. Since 2009 he has been at the Electrical Engineering department, Indian Institute of Technology Bombay as an Assistant Professor. He received a silver medal for the best performance in the Bachelor of technology program of the Electrical Engineering department at the Indian Institute of Technology, Kanpur, India. His current research interests include sampling theory, and statistical and distributed signal processing.
The living, work, and industrial environment of the future will comprise of environments formed by extremely large numbers of devices that are producers and consumers of information, and whose interaction will be facilitated by the Internet of Things (IoT). While such networks and systems will play a critical role in facilitating the automation and improving the efficiency of a number of applications, and enhancing the efficiency of overall quality of life, there are a number of technical challenges in the way of their implementation. The most fundamental of these challenges is that of scale, resulting from the explosion in the number of devices in such networks (expected to be in the trillions according to current estimates). This tutorial will provide an introduction and in depth coverage of the challenges associated with facilitating such large scale communications. The tutorial will start with an introduction to the characteristics of IoT and machine to machine (M2M) communication systems and the outstanding key challenges to be overcome. Next, the tutorial will focus on the network access technologies for IoT and M2M communications, including both the capillary (IEEE 802.15.4e and low power WiFi) and cellular (ETSI M2M and 3GPP LTE-M) components. For each access mechanism, the performance and scalability aspects will be discussed in detail. Finally, ongoing standardization efforts towards the development of protocols for M2M communications will be discussed.
Prof. Biplab Sikdar received the B. Tech degree in electronics and communication engineering from North Eastern Hill University, Shillong, India, M. Tech degree in electrical engineering from Indian Institute of Technology, Kanpur and Ph.D in electrical engineering from Rensselaer Polytechnic Institute, Troy, NY, USA in 1996, 1998 and 2001, respectively. He joined the Department of Electrical, Computer and Systems Engineering of Rensselaer Polytechnic Institute in 2001 as an Assistant Professor. He is currently an Associate Professor in the Department of Electrical and Computer Engineering of National University of Singapore while on leave from Rensselaer Polytechnic Institute. His research interests include wireless MAC protocols, transport protocols, network security and queuing theory. He currently serves as an Associate Editor for the IEEE Transactions on Mobile Computing and has previously served on the editorial board of the IEEE Transactions on Communications. Biplab is a member of Eta Kappa Nu, Tau Beta Pi and a senior member of IEEE.
Lattice codes may be viewed as linear codes designed for use over Gaussian-noise channels, and more generally, channels with real-valued input and output alphabets. It is by now well-established that lattice codes can achieve the capacity of the additive white Gaussian noise (AWGN) channel. Lattice coding schemes also provide the best-known achievable rates in many multi-user communication scenarios. They form the basis of the compute-and-forward strategy that lies at the heart of physical-layer (wireless) network coding. Beyond reliability of communication, lattice-based coding schemes have also been proposed for information-theoretically secure communication (also referred to as physical-layer security).
Lattice codes stand today where algebraic codes were about 25 years back. They have been shown to be theoretically capable of providing the best rates for reliable (and secure) communication in many AWGN settings, but they have struggled to fulfill their potential in practice. The main problem is that we still do not know how to construct encoders and decoders with low implementation complexity for good lattice codes in high dimensions.
In this tutorial, we will give an overview of the theory and applications of lattice codes. We will begin with the necessary mathematical definitions and background for lattices. The earliest application of lattices in the information-theoretic context is in vector quantization, which we will briefly describe. The main focus of the tutorial will be on AWGN channel coding and multi-user communications applications. We will provide a survey of the approaches being used to make lattice coding practical, and list some of the open research problems in this field.
Full-duplex communication (FDC) is a very elegant and lucrative technique for doubling the data rate of a wireless system for the same bandwidth. While most current communication devices (cell phones, wireless routers) appear full-duplex, they are indeed half-duplex systems. In current devices, the transmit and the receive data is separated either in time or frequency.
However, in a ''real'' full-duplex communication system the transmitter sends and receives information in the same time and frequency band thus effectively doubling the data rate. Since, in a FDC system, the transmission and reception happens in the same frequency band at the same time, the transmit data interferes with the receive data. In a typical wireless system the received signal power is about 100 dB less than the transmit signal. Since the transmit signal is known, it can be subtracted (at least theoretically) to recover the received signal. Moreover, this self-interference should be removed in the RF (or the analog baseband) so as to prevent ADC saturation and non-linearities because of the high self-interference power. However, removal of self-interference is complicated by limitations of the RF and analog circuits.
Current wireless systems and standards are not designed for "real" full-duplex systems. Even if an ideal full-duplex node is realized, it is not clear as to how to utilize these nodes in a network.
In this tutorial, we will cover the following topics:
1)Introduction to full duplex wireless communications
2)Basic self-interference problem
- analog, RF issues, challenges in cancellation
- current techniques for cancellation
- research directions
3)Leveraging full-duplex nodes in wireless networks
Prof. Radha Krishna Ganti is an Assistant Professor at the Indian Institute of Technology Madras, Chennai, India. He was a Postdoctoral researcher in the Wireless Networking and Communications Group at UT Austin from 2009-11. He received his B. Tech. and M. Tech. in EE from the Indian Institute of Technology, Madras, and a Masters in Applied Mathematics and a Ph.D. in EE from the University of Notre Dame in 2009. His doctoral work focused on the spatial analysis of interference networks using tools from stochastic geometry. He is a co-author of the monograph Interference in Large Wireless Networks (NOW Publishers, 2008). He received the 2014 IEEE Stephen O. Rice Prize, and the 2014 IEEE Leonard G. Abraham Prize.
Prof. Aniruddhan S. is an Assistant Professor at the Indian Institute of Technology Madras, Chennai, India. He obtained a B. Tech. degree in Electrical Engineering from IIT Madras in 2000. He received his MS and Ph.D. degrees from the University of Washington, Seattle in 2003 and 2006 respectively. Between 2006 and 2011, he worked in the RF-Analog group at Qualcomm Incorporated, San Diego where he designed RF integrated circuits for cellular applications. He is a senior member of the IEEE. His research focusses on analog and RF IC design for communications applications.
In recent years, machine learning has been successfully used to solve a number of challenging problems in image analysis and understanding. In this tutorial, we will discuss:
(a) some of the fundamental ideas related to machine learning
(b) formulating and solving a set of computer vision problems using machine learning
(c) examples of successful methods in computer vision where machine learning has made a difference
Machine learning, in general, deals with the methods that use data and previousexperience to design or improve solutions. In this process, often one solves an appropriate optimization problem to find the best possible solution. Use of data in designing the solution has helped us to come up with solutions that match better with the human expectations.
Though many different ideas from machine learning have been successfully used in image analysis, we limit our focus to three popular directions: (i) Energy Minimization (ii) Support Vector Machines and (iii) Learning image representations. To demonstrate the ideas, we will use problems from low, mid and high level vision.
Prof. C. V. Jawahar is a Professor at IIIT Hyderabad. He works in the broad areas Computer Vision and Machine Learning.
Communication networks have evolved from specialized research and tactical projects to large-scale and highly complex interconnections of intelligent devices, increasingly becoming more commercial, consumer oriented, and heterogeneous. Propelled by emergent social networking services and high-definition streaming platforms, network traffic has grown explosively thanks to the advances in processing speed and storage capacity of state-of-the-art communication technologies. As “netizens” demand a seamless networking experience that entails not only higher speeds but also resilience and robustness to failures and malicious cyber attacks, ample opportunities for signal processing (SP) research arise. The vision is for ubiquitous smart network devices to enable data-driven statistical learning algorithms for distributed, robust, and online network operation and management, adaptable to the dynamically evolving network landscape with minimal need for human intervention.
This tutorial aims to delineate the analytical background and the
relevance of SP tools to network monitoring, introducing the SP audience
to the concept of dynamic network cartography—a framework to construct
maps of the dynamic network state in an efficient and scalable manner
tailored to large-scale heterogeneous networks. Towards this end, the
tutorial will cover the following exciting topics:
1. Prediction of partially observed dynamical processes over networks via dictionary learning
2. Dynamic network delay cartography via Kriged Kalman filter
3. Network distance prediction
4. Dynamic anomalography: unveiling traffic anomalies via sparsity and low-rank
5. RF cartography for cognitive radio networks
Prof. Ketan Rajawat (firstname.lastname@example.org) received his B.Tech and M.Tech degrees in Electrical Engineering from the Indian Institute of Technology (IIT) Kanpur, in 2007, and his Ph.D. degree in Electrical and Computer Engineering from the University of Minnesota in 2012. Currently, he is an assistant professor in the Department of Electrical Engineering, IIT Kanpur. His research interests lie in the areas of Signal Processing and Communication Networks. His current research focuses on cross-layer network optimization, resource allocation, and SP-assisted network monitoring.